Hamilton Smith on The Search for Restriction Enzymes at Hopkins
  Hamilton Smith     Biography    
Recorded: 03 Mar 2006

Yeah, so I went to Hopkins in ’67. Fortunately I was able to—since Mike was actually still on sabbatical at that time – he was on sabbatical I think from July of ‘66 to July of ‘67 – so I was in charge of his lab, and I had a chance to go ahead and apply for an NIH grant, which rambled on and on, but I got a pretty good score. So fortunately I was able to go to Hopkins and have money to start out with. In those days they didn’t provide you with a huge amount of money. You often would just go and teach until you had some money to work with. Well, my grant was for continuing my work in lysogeny. In fact I was heading towards a biochemical analysis of the integration system. I started doing that, and I was lucky to get a graduate student my first couple of months there and I started him working on some things. But then the lambda integration work, because there were fifty labs on lambda, and there were two or three very good people working on integration at NIH— Max Gottesman, I don’t know if Howard Nash was there at the time – probably not. It sounded like, you know, they were pretty far ahead on it, and turned out they probably weren’t, but anyway – I decided I’d go back to recombination. And again, you know I was totally naïve all during this time; I always do what I want to do, that’s my one characteristic. I never bothered much with the rules of how you do it. I used to—back in college I used to take courses that I didn’t have the prerequisites for, and sometimes they wouldn’t let me do it but I’d always find a way to get in. Anyway, so I started—I went across the street to another professor that was working on Haemophilus influenzae, which had a very efficient transformation system. Every cell in the culture could take up four or five molecules and integrate them into the chromosome over a period of half an hour or so. Extremely efficient. And I thought, well, maybe this would be a good system to study recombination, simply because I could take two preparations of DNA, each with a separate antibody resistance marker, transform them in and ask how many double recombinants you get – a very sensitive assay. And if I could do it in the cell, I could certainly add the molecules to an extract and do the same test. If the molecules were recombined, then when I added the DNA to the cells, if the markers were now on the same chromosome, they would transform much more efficiently. So it had a power of about 10^4 or 10^5 of assaying for recombination. So I started doing extracts with Haemophilus, and my graduate student meanwhile was just letting radioactive DNA be taken up by Haemophilus cells and then looking at the fate of the DNA after uptake. One day he used some labeled P22 DNA that we had in the refrigerator and reported that it just disappeared. So I have to say one other thing, though, before this, and that is that a week before that I gave a journal club on Matt Meselson and Bob Yuan’s first description of the type I restriction enzyme, so that was in our minds – the first restriction enzyme. So my student said to me, “Could it be a restriction enzyme?” And I said, “No, I don’t think so.” But then I went home that evening, and thought, oh gosh, you know, we can assay for it very easily. We’ll know the answer in ten minutes. So the next morning we come in—I had been doing some viscometry, naively expecting recombination to make it more viscous or something like that—Anyway, so we set up two viscometers on the bench. One was a control that had Haemophilus extract and Haemophilus DNA, and the other was Haemophilus extract with P22 DNA. We started sucking up the liquid through the capillary and letting it run through and measuring the time it took the solutions to go through. With the—we already knew with the Haemophilus that nothing would happen; it would remain completely viscous. The nucleases—there were essentially no endonucleases in there. But the other one, within the first time point, three or four minutes, was way down, and then it just continued down exponentially. We knew immediately we had a restriction enzyme. Because it was an enzyme that was recognizing—somehow it was reading sequences, knew that this was another DNA of some sort. So, that’s when I started working on—so I dropped the recombination. Yeah. So, my student and I, we spent two months purifying the enzyme. It was the first biochemistry, you know, protein purification we’d ever done. Then he received papers to report to the military, the Vietnam War was on. So he was pulled off to finish off his master’s before he went in. Kent Wilcox. He’s currently a professor at the Medical College of Wisconsin, I think. Although I think he retired recently. So, beyond the purification he was out of the picture on it. And I continued working about a year on it. First off, a month after I came to Hopkins, Bernard Weiss had come from Harvard where he was working with Charlie Richardson on TPOR ligase. And also Richardson had been working on polynucleotide kinase, and they’d been labeling the ends of DNA using gamma-labeled ATP. So Bernie had a tube of this enzyme in his refrigerator and he also had a tube of very hot ATP – he was doing some experiments with the labeled ATP at the time – so I told him I had this enzyme that was probably cleaving somewhere specifically in the sequence. It turned out of course that Meselson’s enzyme did not do that, but I had assumed from reading his paper that it probably did – you know, it recognized a specific sequence and cleaved it. So I assumed the same for our enzyme. So we cleaved DNA and Barry Weiss gave us the protocol and the materials and I labeled it out, and the first result was that it was specific. There was either an A or a G, about equal amounts, on the 5 prime ends of the fragments. Then Bernie told me that Richardson had shown that if you label the ends of molecules and then chew back with—ok, let’s see—I think you hit it with both pancreatic DNase and also exonuclease I. Exonuclease I will chew back and leave just a dinucleotide. So I went downstairs to Paul Englund’s office – he had come from Kornberg’s lab that same year – and he had done the largest preparation of polymerase in history, from six hundred pounds of E. coli, and he had a freezer full of side fractions from this preparation. The ammonium sulfate cuts were very rich in exonuclease I, which you couldn’t buy commercially at the time. So he gave me a couple of bottles of that. I purified the enzyme and used it, and then when we did the experiment the dinucleotides were also specific assuming that we had the central core of the cut side on either side. At that point—and I should say I had to purify all the dinucleotide markers as well, because you couldn’t buy them. That took two or three months, using DEAE chromatography and so on. So what you did was to cleave up salmon sperm DNA with pancreatic DNase and then fractionate the dinucleotides, trinucleotides, tetra, and so on. So I cut out the dinucleotides and ran them through another column and separated into three or four separate dinucleotide sets, which was good enough to determine the sequence of the dinucleotide. Anyway, when we finished that work, I worked about another six months or so looking at the 3 prime end of the DNA, because I wanted to determine whether it was a flush break or whether it was staggered. And the result was that it was an even break. But at that point I couldn’t figure out how to go further on to sequence, so I wrote up a paper and sent it off to JMB. As soon as I sent it out I went back to work. Meanwhile, Tom Kelly had come into my lab, and he started initially just (serving???) nucleases, and then he said, “What are you working on?” And I said, “I’m trying to finish off this sequence here.” He said, “Well that’s much more important than this, why don’t I work on that?” So I said, “Ok, let’s work on it.” Ok, so a few days later Bernie Weiss is walking in the hall and he says to me, I just read in one of the ads in Science that New England Nuclear is selling P33. P33 has an energy one-tenth that of P32, so you can easily separate the two isotopes—So Tom and I went to the blackboard, and I can still vividly see us standing there, and sketched out the procedure. We were going to uniformly label the DNA with P33and then label just the ends with P32, and then cleave the DNA; we could then isolate the ends of the DNA, but since it’s uniformly labeled we would be able to analyze whatever trinucleotides or whatever we got. So we then—and I set up the column in my office, actually, we set up a urea-DEAE column to fractionate the dinucleotides and trinucleotides and so on, after we had done the labeling. We then electrophoretically analyzed those, picked out the spots that were P32-labeled, and then analyzed the nucleotide composition from the P33label, and we were able to finish off the trinucleotide and also even the tetranucleotide, which was degenerate. So we wrote up the paper. I said that when we – I had finished off – actually I was doing all this myself at this point. I had finished off four bases in the center of the recognition site, two on either side of the cut side, and had shown that it was a flush cut. So I had written up the paper and sent it to JMB, and then promptly forgot about it while Kelly and I were working out the rest of the sequence. And just by accident the editor, Charles Thomas, Charlie Thomas, at Harvard, happened to come down to Hopkins to give a lecture. Of course, I knew him from Cold Spring Harbor days and so on, we used to be at the meetings together. He said, “well, Ham, that’s a very nice paper you have there; is it in press?” And I said I haven’t heard a word. He says, you should have reviewed those comments weeks ago. So he said, I’ll go back and check. So when he went back he found it under a stack of papers on his desk and immediately sent everything to me. And I said, well, at this point I want to withdraw this paper, because we’re now going to have a much more complete story and we’re going to have probably two separate papers, one of which I’m the first author and the other Tom Kelly will be the first author. So he said, OK, you have to write to Kendrew and withdraw it. So we did all that, and then he expedited the review of the second paper, of our two other papers, which I think we had sent in in February to him. And they went through very quickly. Of course there was a long lag between acceptance and getting them published. It didn’t come out until I think August of 1970. It’s kind of an interesting story that – very fortuitous, I think, that they had lost that paper, because the impact was so much greater having a completely finished story.

Hamilton Smith is a U.S. microbiologist born Aug. 23, 1931, New York, N.Y. Smith received an A.B. degree in mathematics at the University of California, Berkeley in 1952 and the M.D. degree from Johns Hopkins University in 1956. After six years of clinical work in medicine (1956-1962), he carried out research on Salmonella phage P22 lysogeny at the University of Michigan, Ann Arbor (1962-1967). In 1967, he joined the Microbiology Department at Johns Hopkins.

In 1968, he discovered the first TypeII restriction enzyme (HindII) and determined the sequence of its cleavage site. In, 1978 he was a co-recipient (with D. Nathans and W. Arber) of the Nobel in Medicine for this discovery.

He is currently the Scientific Director Synthetic Biology and Bioenergy Distinguished Professor at the J. Craig Venture Institute in Rockville, Maryland.